statham-arm wrote: Note to reviewers: the thing I'm most uncertain of in this patch is whether I'm using the right strategy for finding out the information about relocation model.
`ParsePICArgs` looked sensible, but then failed an assertion if there was no "effective triple" set, so I dug out what _looked_ like the obvious way to figure one out, and did that. But I'm not _very_ confident of what an "effective triple" is at all, and I don't know for sure that it's right to call `ComputeEffectiveClangTriple(Args)` to make one. So if there's any reason why that will give the wrong answer in a particular situation (e.g. because some information it depends on isn't finalized yet) then please don't assume I have a clever reason for knowing why it's safe after all :slightly_smiling_face: https://github.com/llvm/llvm-project/pull/149132 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
